The published monthly closed-sale series for Marion had its newest month stuck on September through four consecutive Monday checks, and this blog said so each time and published nothing from it. This morning October is there: a median sale price of $219,887 on 98 sales, with a median of 61 days on market.
September, for comparison, was $220,000 on 99 sales in 59 days. So the new month is, to within a rounding error, the old month. What actually changed this week is September.
The new month barely moved
Take the three measures in order. The median sale price went from $220,000 to $219,887 — a difference of $113 on a figure of $220,000, which is to say no difference at all. Sales went from 99 to 98. The median time to sell went from 59 days to 61.
There is no story in that and we are not going to invent one. A month that arrives on top of the month before it is the most common thing a monthly series does.
September moved, though — and by more
Here is what makes this week worth writing about. We keep last week’s copy of this series, so the two can be differenced month by month rather than remembered. Every month before September is identical in both copies. September is not.
| September, as published | Last week | This morning | Difference |
|---|---|---|---|
| Homes sold | 101 | 99 | −2 |
| Median sale price | $219,935 | $220,000 | +$65 |
| Median days on market | 59.062 | 59 | rounded |
The same published series, read seven days apart. No other month on record changed.
Two of the sales that were on September’s record last week are not on it this morning. We do not know why, and the series does not say — records get added, corrected or withdrawn after a month closes and the published figure catches up later. What matters is the size of it: the revision to September is two of the sales it had, and the change from September to October is one. Last month’s correction is bigger than this month’s news.
That is the useful thing to carry out of this post. When a figure for the month just ended appears anywhere — here, in a market report, in a listing presentation — it is provisional, and the only way to know by how much is to have kept the earlier copy and looked. Most people quoting it will not have.
October against the other Octobers
One comparison a single month does support is against the same month in earlier years, which takes the season out of it. This series carries the six Octobers below.
| October | Median sale price | Homes sold | Median days on market |
|---|---|---|---|
| 2021 | $149,200 | 129 | 72 |
| 2022 | $174,900 | 104 | 64 |
| 2023 | $167,000 | 105 | 49 |
| 2024 | $195,000 | 120 | 55 |
| 2025 | $210,500 | 101 | 67 |
| 2026 | $219,887 | 98 | 61 |
The published monthly series, Octobers only.
October 2026 carries the highest October median of the six and the fewest October sales of the six. Against last October specifically: the median is $9,387 higher, which is 4.5%; three fewer homes sold; and six fewer days to sell at the median.
That is a ranking, not a trend, and the table itself shows why. The October median did not rise every year: in 2023 it came in at $167,000 after $174,900 the year before. Six readings with one reversal in them will not tell you where the seventh goes, and nothing on this page claims to.
What the for-sale side did this week
Separately from the sales figures — different houses, different question — this morning’s query returns 193 active listings in the same ZIP code, against 191 a week ago. 16 listings arrived, 14 left, and 177 were on both readings.
Among those 177, 10 cut an asking price and not one raised one. The typical cut was $10,000, the deepest $125,000 and the shallowest $4,900 — so 6% of the carried-over listings moved, in seven days. The typical figure happens to be the same $10,000 that the regional editions of this blog have reported for three weeks running, which is worth a line and not a paragraph.
Nothing else about the carried-over listings changed at all. No listing’s property-type code moved, which earlier editions have caught and treated as a records event rather than a market one. And the days-on-market field advanced by exactly seven days on every one of the 164 carried-over listings that carries it — a clean seven days measured by the records rather than by our calendar.
Of the 193 listings, 116 carry both a bedroom count and a floor area and 67 carry neither, which is how this blog separates houses from land rather than by trusting a type label. Their median asking price is $236,950. A handful of listings sit on codes this blog has not verified and are therefore not described by type at all.
What this does not tell you
It does not tell you Marion prices rose or fell. The median sale price barely moved between the two published months, and a median is the middle of whichever houses happened to sell — not a measure of what any particular house is worth. It does not tell you why September was revised, why October landed flat, or why 10 sellers lowered a price; the records carry figures and not reasons. It does not forecast anything, and it does not treat the sales figures and the for-sale figures as views of one thing, because they are not.
Common questions
What were Marion home sales in October 2026?
The published monthly series now carries October: a median sale price of $219,887, 98 homes sold, and a median of 61 days on market. Against September — $220,000, 99 sales, 59 days — that is a median $113 lower, one fewer sale and two more days. On a median near $220,000, a $113 difference is not a move.
Why did September’s numbers change?
We do not know, and the series does not say. What we can show is that they did. Last week’s copy of the same series put September at 101 sales with a median of $219,935. This morning’s copy puts it at 99 sales with a median of $220,000. Records get added, corrected and withdrawn after a month closes, and the published figure catches up later. The practical lesson is that a just-published month is provisional.
Is $219,887 a high median for Marion?
It is the highest October median of the six Octobers this series carries, and October also brought the fewest sales of those six at 98. But it is not a straight line upward: the October median fell in 2023, to $167,000 from $174,900 the year before. A ranking is not a trend, and this post does not turn it into one.
How many Marion homes are for sale right now, and are prices being cut?
This morning’s query returns 193 active listings, up from 191 a week ago. Of the 177 listings on both readings, 10 cut an asking price and none raised one — a typical cut of $10,000, the deepest $125,000 and the shallowest $4,900. That is 6% of the carried-over listings in seven days.
Do the sales figures and the for-sale figures describe the same houses?
No, and that is why this post keeps them apart. The closed series describes homes that sold in a past month. The active query describes homes that were for sale on the morning it ran. Those are different sets of houses, so no figure from one is compared with, divided by or subtracted from a figure from the other anywhere on this page.
💬 Greg Holthaus on a number that changes after you quote it
“The part of this worth remembering is not October. It is that September moved after it was already printed, and it moved by more than October moved. Anybody can pull a figure for last month and put it in front of a seller; it takes a week of waiting to find out whether it was right. So if you are being shown last month’s median as the reason to price a house a certain way, the fair question is how firm that number is — and on 99 sales in one ZIP code, two of them coming off is enough to notice. What does not move around is the list of houses a buyer at your price is actually choosing between. That is a shorter list and it is the one worth looking at.”
Thinking about selling in Marion? Call or text 618.925.8654 and we will go through what is actually listed at your price, not just what last month averaged.
Where these numbers come from
The sales figures are the published monthly closed-sale series for Marion’s ZIP code, read this morning and read again from last week’s saved copy so the two could be differenced. Each of the three measures is matched by its own label — median sale price, homes sold, median days on market — never by its position in the file, so a reordering upstream would show up as a missing label rather than as silently transposed figures. All three were present in both copies: 60 monthly observations last week, 61 this morning.
The for-sale figures come from local MLS listing data for the same ZIP code, the identical query both weeks, with the query string recorded in the builder rather than described — an earlier edition in this series could not re-run its own stated query because only a description of it had been written down. Listings were matched one to one on a stable record identifier, never on the address text, and the identifier is not printed. No postal-code filter is applied, because the archive this is differenced against contains several extended postal variants and one neighbouring code; filtering to an exact match would ask a narrower question than the comparison needs.
The days-on-market field is its own check on the window: it advanced by exactly seven days on every one of the 164 carried-over listings that carries the field in both readings. 12 carried-over listings have never carried it and are left out of that count rather than counted as zero; none lost it. Earlier editions caught two records misbehaving this way; this week none did.
About these figures: the sales figures above are closed sales from a published monthly series, and the for-sale figures are local MLS listing data read on the morning of November 11, 2026. No figure from the two sources is combined — they describe different sets of houses. A reduction is a change in an asking price, not a sale and not a valuation. No active listing on this page is described as sold, pending or contingent. No individual listing is named, and no address, seller or record identifier appears anywhere here. No reason is given for the revision, for the flat month or for any price cut, because the records carry figures and not reasons, and nothing on this page forecasts anything. Nothing on this page is a loan offer, a rate quote or any statement about financing: no interest rate, no monthly payment and no lender appears in it, and it makes no claim about what anybody can borrow. It names no broker other than Greg, and no other brokerage. Figures describe listings on the morning they were read; a listing can change its price or leave the market the same afternoon. Source note: figures on this page include local MLS data as published by Redfin (redfin.com).
